How to show up in Google AI Mode

To learn how to rank in Google AI Mode, start with what Google itself says: AI Mode splits a question into subtopics, searches each one, and grounds its answer in pages from the Search index. There is no separate checklist. A page must be indexed and eligible for a snippet, and it earns its place by answering the related sub-questions a fan-out search will ask.

What is Google AI Mode, and how does it find pages?

AI Mode is a search experience that answers in prose and lets you keep asking follow-up questions, with links out to the web. Google's AI Mode help page describes it as expanding what AI Overviews can do "with more advanced reasoning and ways of interacting", and says it "divides your question into subtopics and searches for each one simultaneously."

The important part for a site owner is where the answer comes from. The same help page says AI Mode "relies on Google Search's deep understanding of web information", so responses are supported by web content. Google's launch post for AI Mode adds that it "is rooted in our core quality and ranking systems." In other words, AI Mode does not read some separate copy of the web. It reads the Search index, through the same systems that decide ordinary rankings.

What is query fan-out, and why does it matter for AI Mode SEO?

Query fan-out is Google's name for how AI Mode researches a question: it breaks one query into several related searches, runs them together, and stitches the results into one answer. The launch post says AI Mode "uses a “query fan-out” technique, issuing multiple related searches concurrently across subtopics and multiple data sources and then brings those results together." Google's AI features documentation says both AI Overviews and AI Mode may use the technique.

Here is why that changes how you write. A traditional search matches one query against one ranking. A fan-out search turns a single question into a cluster of sub-questions. Take someone asking AI Mode which accounting software suits a small consultancy that bills in two currencies. The underlying searches might cover multi-currency invoicing, pricing and tax handling, each a separate chance for a page to be pulled in.

So a page that only targets the headline phrase competes for one slot. A page that cleanly answers the sub-questions around that phrase can be relevant to several of the searches the fan-out runs. This is our reading of the mechanism Google describes, not a promise from Google about how pages are chosen, but it is the most useful way we have found to plan content for AI mode SEO.

How to rank in Google AI Mode: what Google says you need

The short answer is that Google lists no AI Mode-specific requirements. Its AI features page says "the best practices for SEO remain relevant" and that "there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." The same page sets the eligibility bar: a page "must be indexed and eligible to be shown in Google Search with a snippet, fulfilling the Search technical requirements."

Google's guide to optimizing for generative AI features adds one more condition worth checking: "a site must be included in Search generative AI features in Search Console to be eligible for display in generative AI features on Google Search." It also states plainly that, from Google Search's perspective, optimising for generative AI search "is optimizing for the search experience, and thus still SEO."

The crawl row is the one we would underline. Google's AI features page lists "ensuring that crawling is allowed in robots.txt, and by any CDN or hosting infrastructure." A firewall rule or CDN bot setting can block a crawler while robots.txt says everything is fine, so checking robots.txt alone tells you only half the story.

What does Google say you do not need for Google AI Mode?

Google is unusually direct here: you do not need new files, special markup or an AI-specific schema type. The AI features page says "you don't need to create new machine readable files, AI text files, or markup to appear in these features" and that there is "no special schema.org structured data that you need to add."

The optimisation guide goes further on llms.txt. It says creating one "will neither harm nor help your site's visibility or rankings in Google Search, as Google Search ignores them." It also lists overfocusing on structured data as a mistake, while recommending you keep using it as part of your overall SEO strategy because it helps with eligibility for rich results.

How to rank in Google AI Mode with sub-question coverage

Build each important page around the main question and the follow-ups a fan-out search would plausibly ask, answered in self-contained passages. This is the procedure we use, sitting on top of the requirements above rather than replacing them.

1. Confirm eligibility first. Check that the page is indexed, can show a snippet, returns a 200 to Googlebot through your CDN and firewall, and that the site is included in generative AI features in Search Console.
2. Write down the parent question. One sentence, the way a customer would ask it, not the keyword.
3. List the sub-questions. What would someone need to know to act on the answer? Cost mechanisms, comparisons, requirements, risks, how long it takes, who it suits. Use Search Console queries, sales calls and support tickets as raw material.
4. Give each sub-question its own heading or block. Phrase the heading as the question, and answer it in the first one or two sentences beneath it.
5. Make each answer stand alone. Name the subject instead of saying "it" or "this approach", so a passage still makes sense when lifted out of the page.
6. Add comparisons where readers compare. Google's launch post says AI Mode is particularly helpful for questions that need "further exploration, comparisons and reasoning." A clear table answers a comparison sub-question in one place.
7. Link the cluster together. Where a sub-question deserves its own page, write it and link both ways, so the fan-out has more than one strong page from you to find.
8. Keep it current. Grounding relies on retrieving "relevant, up-to-date web pages", in the optimisation guide's words, so stale figures and dead advice are a liability.

If you want to see this applied to Google's other AI surface, our post on getting into AI Overviews covers the overlap, and how to optimise for AI search widens it to ChatGPT, Perplexity and Claude.

What we have seen when sites fail the basics

Before any clever optimisation, we check access and structure, because that is where our audits have turned up real problems. On one SEO Autopilot engagement, a US healthcare analytics company on WordPress behind a firewall, the firewall returned 403 to OpenAI's and Anthropic's search crawlers on every page. The trigger was a generic bad-bot rule matching the substring "searchbot". The client's monitoring tool reported all sixteen AI crawlers as accessible, because it read robots rules instead of sending requests.

Those were not Google's crawlers, but the lesson carries straight over to the crawl requirement in Google's docs: test with real requests, not just by reading robots.txt. The same site had 21 of 39 pages with no H1, because the post template rendered the title as an H2. One template change took pages with exactly one H1 from 9 of 39 to 30. Structure problems like that sit underneath every sub-question you write.

Aglocom's free AI visibility checker is a quick first look. It reads robots.txt rules for 11 AI crawlers and requests the homepage once with each crawler's user agent. That is the checker's own request, not the real crawler's, and it covers the homepage only, so treat it as a prompt for a deeper look rather than proof that a crawler can reach your site.

How do you measure Google AI Mode visibility?

Start in Search Console. Google's AI features page says sites appearing in AI features, including AI Mode, "are included in the overall search traffic in Search Console." That means AI Mode clicks and impressions are counted with the rest of your search traffic rather than in a separate report.

In practice we watch three things: impressions and clicks on the pages we built around question clusters, the spread of queries each of those pages earns (a widening spread suggests the sub-answers are being matched), and manual checks of AI Mode for the parent questions that matter commercially. None of that proves why a page was cited, and Google's help page itself says AI Mode "doesn't always get it right", so we treat it as direction rather than a scoreboard.

Patience matters too. The AI features page notes that crawling "can take anywhere from several days to several months, depending on how often our systems determine a page needs to be refreshed." A rewrite will not show up the next morning. If you would rather have someone audit this for you, our GEO and AEO audit checks crawl access, indexing, structure and question coverage page by page.

Frequently Asked Questions